<p>The resolution of near-surface seismic reflection data is often limited by attenuation and scattering in the shallow subsurface which reduces the high frequencies in the data. Compensating for attenuation and scattering, as well as removing the propagating source wavelet in a time-variant manner can be used to improve the resolution. Here we investigate continuous non-stationary iterative time-domain deconvolution (CNS-ITD), where the seismic wavelet is allowed to vary along the seismic trace. The propagating seismic wavelet is then a combination of the source wavelet and the effects of attenuation and scattering effects, and can be estimated in a data-driven manner by performing a Gabor decomposition of the data. For each Gabor window, the autocorrelation is estimated and windowed about zero lag to estimate the propagating wavelet. Using the matrix-vector equations, the estimated propagating wavelets are assigned to the related columns of a seismic wavelet matrix, and these are then interpolated to the time location where the maximum of the envelope of the trace occurs within the iterative time-domain deconvolution. Advantages of using this data-driven, time-varying approach include not requiring prior knowledge of the attenuation and scattering structure and allowing for the sparse estimation of the reflectivity within the iterative deconvolution. We first apply CNS-ITD to synthetic data with a time-varying attenuation, where the method successfully identified the reflectors and increased the resolution of the data. We then applied CNS-ITD to two observed shallow seismic reflection datasets where improved resolution was obtained. </p>

The mode of tectonics by which early Archean proto-continents were deformed was investigated in the Pilbara Craton in Western Australia, which has not been substantially tectonically deformed since ~3.2 Ga. The craton consists of a unique dome and keel structure where vertical, low-grade metamorphism basaltic greenstone keels surround large granitic (TTG) domes. The dominant model for 3.5-3.2 Ga deformation in the Pilbara is gravity-driven vertical tectonics, or partial convective overturn in a hot crust. In this model, the granitic bodies rose upward as solid-state diapirs, and the greenstones "sagducted" downward around the granitic bodies. Australian scientists acquired deep seismic reflection data crossing a granitoid-greenstone boundary. Their processing did not image the geologically mapped steep dip of the boundary because standard methods limit the maximum dip. A 37-km section of these data were reprocessed using 2D Kirchhoff prestack depth migration to include vertical dips. The western half of the migrated section images a granitoid dome with weak to no reflectivity that extends deeper than 4 km. The eastern half images 2-3 km of layered volcanic rocks of the Fortescue Group overlying the greenstones. Seismic velocity models created using travel-time tomography suggest a thin weathering layer overlying slightly fractured crystalline rocks. These fractures close within 200-300 m depth, and velocity reaches bedrock speeds consistent with expected values of granitoids to the west and volcanic rocks of the Fortescue Group to the east. The best migrated image contains several reflections with dips (~45-55˚) cross-cutting each other from both directions at the location of the expected granitoid-greenstone boundary. This strongly suggests the presence of steep dips in the upper ~1.5 km but does not provide a definitive image. This inconclusive result is due to strong surface-wave noise, the crooked 2D seismic line, and the 3D nature of the geologic boundary at the seismic line. A very small seismic velocity gradient within the crystalline bedrock limits the maximum depth to which vertical features can be imaged. / Master of Science / The Pilbara craton is one of the few exposed and intact pieces of continents that were formed ~3.2 billion years ago. This research analyzes how these early land masses were deformed. There are two methods by which early land masses evolved—vertical tectonics (a more rudimentary, gravity-driven form of plate movement) or horizontal tectonics (which is closer to modern-day tectonics and requires many stages of deformation). This area has a unique dome-and-keel structure where greenstones (metamorphosed volcanics) are vertically wrapped around large granitic domes. Studying the vertical features of the greenstones will allow us to ascertain how tectonics evolved in the area. A seismic survey was conducted in 2018 in the area. These data were reprocessed to include steep dips to extract the exact location of the steeply dipping boundary between the dome and keel structure at depth. The resulting image contains inconclusive evidence due to the physical limitations of the geology and the sharp bend in the seismic line. Further studies need to be done to determine if the Pilbara Craton was formed by vertical tectonics.

The use of seismic methods in hard rock environments in Western Australia for mineral exploration is a new and burgeoning technology. Traditionally, mineral exploration has relied upon potential field methods and surface prospecting to reveal shallow targets for economic exploitation. These methods have been and will continue to be effective but lack lateral and depth resolution needed to image deeper mineral deposits for targeted mining. With global need for minerals, and gold in particular, increasing in demand, and with shallower targets harder to find, new methods to uncover deeper mineral reserves are needed. Seismic reflection imaging, hard rock borehole data analysis, seismic inversion and seismic attribute analysis all give the spatial and volumetric exploration techniques the mineral industry can use to reveal high value deeper mineral targets. / In 2002, two high resolution seismic lines, the East Victory and Intrepid, were acquired along with sonic logging, to assess the feasibility of seismic imaging and rock characterisation at the St. Ives gold camp in Western Australia. An innovative research project was undertaken combining seismic processing, rock characterization, reflection calibration, seismic inversion and seismic attribute analysis to show that volumetric predictions of rock type and gold-content may be viable in hard rock environments. Accurate seismic imaging and reflection identification proved to be challenging but achievable task in the all-out hard rock environment of the Yilgarn craton. Accurate results were confounded by crocked seismic line acquisition, low signal-to-noise ratio, regolith distortions, small elastic property variations in the rock, and a limited volume of sonic logging. Each of these challenges, however, did have a systematic solution which allowed for accurate results to be achieved. / Seismic imaging was successfully completed on both the East Victory and Intrepid data sets revealing complex structures in the Earth as shallow as 100 metres to as deep as 3000 metres. The successful imaging required homogenization of the regolith to eliminate regolith travel-time distortions and accurate constant velocity analysis for reflection focusing using migration. Verification of the high amplitude reflections within each image was achieved through integration of surface geological and underground mine data as well as calibration with log derived synthetic seismograms. The most accurate imaging results were ultimately achieved on the East Victory line which had good signal-to-noise ratio and close-to-straight data acquisition direction compared to the more crooked Intrepid seismic line. / The sonic logs from both the East Victory and Intrepid seismic lines were comprehensively analysed by re-sampling and separating the data based on rock type, structure type, alteration type, and Au assay. Cross plotting of the log data revealed statistically accurate separation between harder and softer rocks, as well as sheared and un-sheared rock, were possible based solely on compressional-wave, shear-wave, density, acoustic and elastic impedance. These results were used successfully to derive empirical relationships between seismic attributes and geology. Calibrations of the logs and seismic data provided proof that reflections, especially high-amplitude reflections, correlated well with certain rock properties as expected from the sonic data, including high gold content sheared zones. The correlation value, however, varied with signal-to-noise ratio and crookedness of the seismic line. Subsequent numerical modelling confirmed that separating soft from hard rocks can be based on both general reflectivity pattern and impedance contrasts. / Indeed impedance inversions on the calibrated seismic and sonic data produced reliable volumetric separations between harder rocks (basalt and dolerite) and softer rock (intermediate intrusive, mafic, and volcaniclastic). Acoustic impedance inversions produced the most statistically valid volumetric predictions with the simultaneous use of acoustic and elastic inversions producing stable separation of softer and harder rocks zones. Similarly, Lambda-Mu-Rho inversions showed good separations between softer and harder rock zones. With high gold content rock associated more with “softer” hard rocks and sheared zones, these volumetric inversion provide valuable information for targeted mining. The geostatistical method applied to attribute analysis, however, was highly ambiguous due to low correlations and thus produced overly generalized predictions. Overall reliability of the seismic inversion results were based on quality and quantity of sonic data leaving the East Victory data set, again with superior results as compared to the Intrepid data set. / In general, detailed processing and analysis of the 2D seismic data and the study of the relationship between the recorded wave-field and rock properties measured from borehole logs, core samples and open cut mining, revealed that positive correlations can be developed between the two. The results of rigorous research show that rock characterization using seismic methodology will greatly benefit the mineral industry.